Analysis of Vancomycin Dispatch Times

Brian Gulbis

March 25, 2017

Distribution

In all of the following figures, values where the scheduled collection time occured > 1 hour prior to the order time were removed (n = 44). It is unclear exactly why the the scheduled collection time for these orders was back-dated to be prior to the time of ordering.

Distribution of time from task dispatch to scheduled collection

It is clear that Stat orders dispatch immediately, but Routine and Timed Study are much more variable, with most dispatching 1-3 hours before the scheduled collection time.

Distribution of time from task dispatch to scheduled collection

Summary measures for the time from dispatch to scheduled collection

type category minimum q1 median mean q3 maximum na
All All -3.017 0.000 0.783 1.709 1.783 33.550 33
Future All -2.250 0.783 1.767 2.577 2.283 33.550 16
Future Routine -2.250 0.250 1.117 2.607 3.242 23.800 6
Future Stat -0.083 2.000 4.867 7.218 11.150 24.280 NA
Future Timed Study 0.017 0.879 1.767 2.272 2.267 33.550 10
Now All -3.017 -0.017 -0.017 -0.026 0.000 0.917 17
Now Routine -3.017 -0.800 -0.333 -0.502 -0.033 0.817 6
Now Stat -0.800 -0.017 -0.017 0.000 0.000 0.917 7
Now Timed Study -0.483 0.000 0.217 0.292 0.633 0.917 4

Future Orders

Distribution of time from task dispatch to scheduled collection by day of week

Day of week and hour of day are based on the scheduled collection time.

Distribution of time from task dispatch to scheduled collection by hour of day

Distribution of time from task order to dispatch

Distribution of time from task order to dispatch

Control Charts

Time from dispatch until scheduled collection for Future Routine orders

Time from dispatch until scheduled collection for Future Routine orders

Time from dispatch until scheduled collection for Future Stat orders

Time from dispatch until scheduled collection for Future Stat orders

Time from dispatch until scheduled collection for Future Timed Study orders

Time from dispatch until scheduled collection for Future Timed Study orders

Correlation Assessment

Time between order to scheduled collection and order to dispatch for future orders

There is a strong correlation between the time from order to scheduled collection and the time from order to dispatch.

Correlation between the time from order to scheduled collection and the time from order to dispatch

type category estimate statistic p.value parameter conf.low conf.high method alternative
All All 0.991 410.997 0.000 3224 0.990 0.991 Pearson’s product-moment correlation two.sided
Future All 0.991 339.529 0.000 2149 0.990 0.992 Pearson’s product-moment correlation two.sided
Future Routine 0.996 289.329 0.000 640 0.996 0.997 Pearson’s product-moment correlation two.sided
Future Stat 0.277 2.692 0.009 87 0.073 0.459 Pearson’s product-moment correlation two.sided
Future Timed Study 0.972 154.543 0.000 1418 0.969 0.974 Pearson’s product-moment correlation two.sided
Now All -0.008 -0.276 0.783 1073 -0.068 0.051 Pearson’s product-moment correlation two.sided
Now Routine 0.034 0.408 0.684 143 -0.130 0.196 Pearson’s product-moment correlation two.sided
Now Stat 0.134 3.766 0.000 775 0.064 0.202 Pearson’s product-moment correlation two.sided
Now Timed Study 0.264 3.363 0.001 151 0.110 0.406 Pearson’s product-moment correlation two.sided

Time between order to scheduled collection and dispatch to scheduled collection

There is little correlation between the time from order to scheduled collection and time from dispatch to scheduled collection. This suggests that the dispatch time is triggered by the scheduled collection time, regardless of how far in advance the lab is ordered, with the exception of Stat orders which dispatch immediately.

Correlation between the time from order to scheduled collection and time from dispatch to scheduled collection

type category estimate statistic p.value parameter conf.low conf.high method alternative
All All 0.238 13.907 0.000 3224 0.205 0.270 Pearson’s product-moment correlation two.sided
Future All 0.115 5.381 0.000 2149 0.073 0.157 Pearson’s product-moment correlation two.sided
Future Routine 0.126 3.206 0.001 640 0.049 0.201 Pearson’s product-moment correlation two.sided
Future Stat 0.921 22.059 0.000 87 0.882 0.948 Pearson’s product-moment correlation two.sided
Future Timed Study 0.182 6.950 0.000 1418 0.131 0.231 Pearson’s product-moment correlation two.sided
Now All 0.613 25.396 0.000 1073 0.574 0.649 Pearson’s product-moment correlation two.sided
Now Routine 0.290 3.623 0.000 143 0.133 0.433 Pearson’s product-moment correlation two.sided
Now Stat 0.968 106.808 0.000 775 0.963 0.972 Pearson’s product-moment correlation two.sided
Now Timed Study 0.987 74.553 0.000 151 0.982 0.990 Pearson’s product-moment correlation two.sided

Time between dispatch to scheduled collection and scheduled to actual collection

There is no correlation between the amount of time prior to scheduled collection that the task dispatched and when the lab was actually collected.

Correlation between the amount of time prior to scheduled collection that the task dispatched and when the lab was actually collected

type category estimate statistic p.value parameter conf.low conf.high method alternative
All All -0.079 -4.210 0.000 2804 -0.116 -0.042 Pearson’s product-moment correlation two.sided
Future All -0.103 -4.405 0.000 1827 -0.148 -0.057 Pearson’s product-moment correlation two.sided
Future Routine -0.049 -1.180 0.238 578 -0.130 0.033 Pearson’s product-moment correlation two.sided
Future Stat 0.151 1.373 0.173 81 -0.067 0.355 Pearson’s product-moment correlation two.sided
Future Timed Study -0.170 -5.870 0.000 1164 -0.225 -0.113 Pearson’s product-moment correlation two.sided
Now All -0.251 -8.081 0.000 975 -0.308 -0.191 Pearson’s product-moment correlation two.sided
Now Routine -0.285 -3.342 0.001 126 -0.437 -0.118 Pearson’s product-moment correlation two.sided
Now Stat -0.072 -1.943 0.052 718 -0.145 0.001 Pearson’s product-moment correlation two.sided
Now Timed Study -0.094 -1.067 0.288 127 -0.263 0.080 Pearson’s product-moment correlation two.sided

Back-Dated Orders

Provider role that ordered back-dated levels

Order type used for back-dated levels

Future Orders

Orders for future labs by each order priority. Future labs are considered those > 1 hour from the time of ordering.

A large number of Routine orders (81%) were actually intended as future timed orders.

Provider role placing future orders using Routine or Stat

Provider role placing Now orders using Timed Study